Back to catalog
DataAdvanced
Production Data Science - From EDA to Deployment
Beautiful notebooks are not enough. This course teaches you how to turn machine learning experiments into maintainable systems with DVC, MLflow, FastAPI serving, CI/CD for ML, and production drift monitoring.
16 lessonsCertificate includedUSD 10 (~ARS 10.000)
Course syllabus
1From notebook to production code
3 lessons
From notebook to production code
- Problems with notebooks in production
- Refactoring ML code: modules and pipelines
- Versioning data and experiments with DVC
2MLflow and experiment management
3 lessons
MLflow and experiment management
- Experiment tracking with MLflow
- Model Registry: versions and stages
- Comparing and selecting models
3Model serving
3 lessons
Model serving
- FastAPI APIs for ML models
- Serialization: joblib, pickle, and ONNX
- Containerizing a model for production
4CI/CD for ML
3 lessons
CI/CD for ML
- Automated training pipelines
- Model testing: unit tests and validation
- GitHub Actions for ML pipelines
5Production monitoring
3 lessons
Production monitoring
- Data drift and model drift: detection
- Evidently AI: data quality reports
- Alerts and automatic retraining
6Final project
1 lessons
Final project
- Full pipeline: data -> model -> API -> monitoring deployed in the cloud
What you will learn
Machine learningMLflowDVCFastAPIDockerCI/CD for MLEvidently AI
Certificate
Data Scientist Certificate - CumbreAcademy
Ready to start?
Investment: USD 10 (~ARS 10.000)
Buy accessWant access to every course?
Total Access gives you this course and all the others for $20/month.
This course: USD 10 (~ARS 10.000) - Total Access: $20 USD/month (all courses)
See Total Access